Mice homozygous for lpr (a defect at the fas locus) and gld develop autoimmunity and profound lymphadenopathy characterized by the accumulation of two functionally anergic T cell subsets, a predominant B220+CD4-CD8- (DN) population and a minor B220+CD4+ population. In the past year we examined the cellular requirements for the accumulation of DN T cells and the nature of the block in signal transduction in these cells. In addition, lpr and gld mice were examined for evidence of abnormalities in T cell deletion in the periphery that may relate to the defective expression of the fas receptor. Depletion of CD8+ T cells in vivo and ongoing adoptive transfer of gld lymphocyte subsets into scid mice revealed that: 1) CD8+ T cells are required for the accumulation of DN T cells; 2) the precursors of DN T cells are present in gld BM, LN and spleen; 3) CD4+ and CD4+B220+ T cells do not spontaneously differentiate into DN T cells; and 4) DN T cells do not grow autonomously in scid mice or differentiate into CD4+ or CD8+ T cells. Treatment of mice with mAb to IFN-gamma and TNF-alpha had no effect on the accumulation of DN T cells but greatly reduced the levels of serum anti-ds DNA Ab. Signal transduction studies showed that DN T cells proliferated and secreted IL-2 in response to TCR crosslinking and Ab- mediated ligation of CD28. This response differed quantitatively and qualitatively from that of normal T cells, suggesting that DN T cells have a higher than normal threshold for stimulation but potentially may be able to proliferate in vivo. Stimulation of lpr and gld mice with SEB confirmed that DN T cells can proliferate in vivo, albeit less efficiently than CD4+ or CD8+ T cells. All T cell subsets expressing V-beta-8 were deleted after SEB treatment but the deletion of lpr and gld CD8+ T cells was impaired. On the basis of these and earlier studies, we propose that DN T cells arise from peripheral, previously activated CD4+, CD8+ or CD4+CD8+ T cells that have down-regulated the expression of CD4, CD8 and the TCR and normally would be deleted by a fas-dependent mechanism. We further propose that the fas receptor does not play a major role in the deletion of self or foreign super Ag- reactive T cells in the thymus or in the periphery.